Eko.ai aims to disrupt cardiologists with AI

Yet the turn of events made him realise how inefficient the existing heart screening process is. And being the en­trepreneur that he is, he saw a commercial opportunity to be exploited.

As Hare explains, ultrasound is the most common meth­od to diagnose a patient’s heart. This procedure, which is known as echocardiography (echo), creates two- or three-dimensional images of the heart. It allows doctors to look for signs of heart disease, such as weak heart muscle, blood clots inside the heart and poorly functioning heart valves.

Getting an echo, however, is not as easy as walking into your neighbourhood clinic and requesting for a heart check-up. The procedure is usually available only in hospitals and medical centres because these health institutions and cen­tres boast sonographers and cardiologists, who are special­ly trained to conduct an echo and analyse the resulting im­ages. General physicians at clinics do not have the required expertise or equipment to do so.

But what if you could make the screening process easi­er and convenient? After all, there are now transducers that can be attached to a smartphone, thus allowing self-diagno­sis anytime, anywhere. And the analysis could be generated by artificial intelligence (AI) machines, says Hare.

According to him, there are plenty of unused ultrasound images produced from many echo procedures over the years. Of the 70 to 90 images generated from each procedure, only a fraction is actually used to diagnose a patient’s heart con­dition, he says. The rest of the images get archived.

Hare says AI machines can be trained using the unused collection of ultrasound images to correctly diagnose a pa­tient’s heart condition. In particular, they can learn how to correctly identify characteristics from the ultrasound imag­es of patients that correlate to heart diseases.

In 2017, Hare came out of retirement to co-found Eko.ai together with his wife Carolyn Lam, who ironically is a car­diologist. The start-up aims to democratise the diagnosis of heart diseases using AI — even if it means potentially put­ting his wife out of her full-time job.

Lam is currently the director of the Clinical & Transla­tional Research Office at the National Heart Centre of Sin­gapore. She is also the Professor of Cardiology at Duke-NUS Graduate Medical School.

Eko.ai has two other co-founders: Yoran Hummel and Paul Seekings. Hummel is a cardiovascular researcher specialis­ing in echocardiography, diastology and cardio-mechanics. He is also founder and general manager of Groningen Imag­ing Core Lab. Seekings is a former chief technology officer at Embodied Sensing, an Internet of Things solutions provider.

Not just clinical, but research too

Thus far, Eko.ai has yet to obtain any regulatory approval for its AI diagnostic software. But it is hoping to do so, starting with the US Food and Drug Administration this year. Follow­ing that, it will pursue a CE certification in Europe and an approval from Singapore’s Health Sciences Authority (HSA).

Hare says this approach seems to be the common route taken by medical equipment manufacturers. “[The HSA] tends to like it when it’s approved of us elsewhere [first],” he tells The Edge Singapore in an interview. Once that is done, Eko.ai will be able to market its AI diagnostic software in Singapore, the US and Europe.

While Eko.ai intends to sell its AI diagnostic software as a clinical tool, the company is also interested to market it as a research tool for pharmaceutical companies and drug manufacturers. In particular, the AI diagnostic software can be used as an echo database for the pharmaceutical compa­nies and drug manufacturers to identify the right patients to test and to develop new drugs.

“When they do a trial to test a drug, they always use echoes as part [of the process]. It’s a standard thing,” says Hare. “Any tool that can help them use echoes to better iden­tify the correct patients to put in the trial as a better identi­fier, classifier and quantifier [would be great]. And the fact that it can do things automatically, it’s a no-brainer for them to bring such efficiencies into the process.”

According to Hare, the market potential is huge and the sales process is easier, given that the pool of potential cus­tomers is concentrated. As such, securing one or two of them would be a lucrative catch. “The downstream diagnostics [segment] is literally every doctor around the world. But in the upstream segment, it is a dozen [large] pharmaceutical companies,” he says.

“It’s a big, big market,” echoed Lam, who was at the same interview. She notes that Eko.ai has already set up a network of centres that have archived these echoes.

To be sure, Eko.ai is not the only company trying to use AI as a diagnostic tool for heart diseases. But Hare says he is not afraid of the competition as everyone has a different, albeit similar, offering. For instance, he notes that there is a US start-up that has a similar offering that targets only the diagnostic segment. “What we’re trying to do is provide a complete solution both to pharmaceutical companies and doctors,” he says.

Hare is also unfazed by the competition from multina­tional corporations, such as General Electric and Philips, that are developing their own AI diagnostic software. He reckons that Eko.ai has a superior solution compared to them. “We’ve had a number of conversations with them, and I be­lieve we’re quite well-advanced in a lot of stuff [compared to them],” he says.

Investors are already backing Eko.ai. Early this year, the start-up raised US$4 million ($5.7 million) in a fund­ing round co-led by Singapore government-linked EDBI and venture capital firm Sequoia India. Partech Ventures, SG In­novate and Startup Health Transformer also participated in the round. The start-up previously received financial support from the ATPL Gap funding programme under the Agency for Science, Technology and Research (A*Star).

Last year, Eko.ai won the Startup SG grand prize of $250,000 at the Slingshot 2019 start-up competition, which took place during the Singapore FinTech Festival. It was also one of two sector winners for the HealthTech, BioTech and MedTech category, enabling the start-up to pocket an addi­tional $50,000 Startup SG grant.

Back to building a business

In a way, Hare has come full circle with the founding of Eko.ai. Prior to Eko.ai, Hare had co-founded travel agency eDreams in 1999. The company was headquartered in Barcelo­na and became Spain’s top online travel agency. Between 2004 and 2007, the agency saw tremendous growth across Europe.

eDreams’ spectacular growth attracted the attention of private equity firms. In 2006, the US-based TA Associates acquired the company for EUR153 million ($235.6 million). In 2010, European investment fund Permira became its ma­jority shareholder.

Following that, eDreams merged with Go Voyages, one of its rivals. It also acquired two other travel brands, Opo­do and Travellink. As a result of these transactions, the en­larged entity was rebranded as eDreams Odigeo. The com­pany became Europe’s top online travel agency and ranked among the top five globally.

In 2014, eDreams Odigeo made a significant milestone. Shares of the company were listed on the Madrid Stock Ex­change at an IPO price of EUR10.25. The stock subsequently rose to as high as EUR11.40. Hare, who had served as president and independent director at the company, resigned in 2015.

Since then, Hare had been busy making angel invest­ments in start-ups that are involved in machine learning, digital platforms and e-commerce. “I love scouting for tech businesses,” he says.

According to investment site Investor’s Globe, Hare in­vested in Chariot, a New York-based public transportation company that was aiming at disrupting driving and ride-sharing. Chariot was acquired by Ford Motor in 2016. His other investments include blockchain and data security company Blockstream, e-commerce company Dagne Do­ver, IT healthcare company AirCare and online wholesale e-commerce company Boxed.

Now that Hare is back to building a business, looking ahead, where does he envision Eko.ai to be? The big dream is to give people more control over their health, he says. “I was one of these guys, who constantly get their blood test­ed and etcetera,” he says. “But I knew nothing about my heart — the thing that is mostly likely to kill you … So I would love for everybody to just have full control of their heart health, to be able to check it.”